Last Updated: 14/03/2025
Spatio-temporal modelling of climate variability and malaria transmission across time scales to assess causality, improve forecasting and strengthen early warning
Objectives
The overarching goal of this project is to deepen current understanding of the effects of climate change on the burden of malaria by innovating methods that address the non-linear interactions of the malaria drivers on the transmission dynamics and the non-stationarity of the data. The specific objectives are to:
- assess time-delayed causal effects of climatic and non-climatic factors on the changes of malaria incidence across different time scales and levels of transmission;
- develop age-structured stochastic metapopulation malaria transmission models which incorporate climate and control intervention effects;
- develop non-stationary model formulations for short and medium-term malaria forecasts taking into account climate variation across time-scales; and
- evaluate the performance of the new tools on common datasets and their computational feasibility of implementation within a model-based early warning system.
Model-based malaria surveillance that incorporates climate effects is recognised as an adaptation strategy to address the impacts of climate variability to malaria outbreaks. Climate is one of the drivers of transmission, however other factors such as control interventions and socio-economic development can influence the disease dynamics. During the first phase of the project, statistical and mathematical models were developed to quantify the contribution of climatic and non-climatic factors to malaria and evaluate the added value of mathematical transmission models to outbreak forecasting compared to statistical models. Data from HDSS in Kisumu showed that temperature has a protective effect similar to the effect of bednet, but this is opposed by rainfall. These relations varied across seasonal scales. The high correlations between climatic factors and non-linear relations to malaria made it difficult to have always consistent results from statistical models which are able to estimate associations but neither identify causal relations nor take into account non-linear interactions of the malaria drivers. The non-stationarity of the malaria data and varying effect of predictors across seasons suggested that forecasting models should perform better when non-stationarity is relaxed. The proposed project for the second phase builds on the findings, insights and research gaps that have been identified. The objectives will be accomplished by (a) employing and further developing methods to assess causality in time series data and to forecast coupling wavelets with machine learning and dynamic, time delay embedding models; (b) using powerful computational algorithms such as iterated filtering, particle Markov chain Monte Carlo (pMCMC) and Hamiltonian Monte Carlo (HMC) simulation; and (c) analysing existing data from the HDSS in Nouna (Burkina Faso) and Kisumu (Kenya), the Health Information System in Burkina Faso and Kenya, outputs of downscaled climate models, high resolution hydrometeorological data, satellite-based climatic products and other gridded environmental proxies.
Jan 2024 — Dec 2026
$554,790
Related Resources
